5fa1a76
1
2
3
4
5
) width, height = image.size However, this model includes a brand new [~transformers.LayoutLMv2Processor] which can be used to directly prepare data for the model (including applying OCR under the hood).
